Dynamic Response And Surface Roughness Of Large - Scale Holes Under Multiple Variable Constraints

Hard whirling for large-scale thread with the features of workpiece multi-point variable floating support and servo press device dynamic clamping constraints and milling cutter long distance movement and more tools periodic intermittent gradual cutting makes the dynamic response and cutting characteristics of workpiece system and the surface roughness of thread raceway and so on obviously different from common cutting.To realize the optimizing control of the whirling process and improve the total length quality of large-scale thread in hard whirling further, the CNC whirling machine for large-scale thread taken as the research object,the dynamic response of workpiece system in whirling process and surface roughness of thread raceway have been researched respectively.Firstly the process parameters experiment for whirling was carry out, the typical cutting force and vibration and surface profile morphology of thread raceway were analyzed and their characteristic values were extracted to research the influence law and degree of the process parameters and muti variable supporting and clamping constraints on dynamic response of workpiece system and surface roughness of thread raceway and the optimal cutting parameters combinations were proposed.Then according to the Rayleigh Beam theory and GAMM the dynamic model for large-scale thread workpiece system was set up in which the rotation of workpiece and the chunk-top and muti variable supporting and clamping constraints condition and the the cutting force moving along the axial direction were taken into consideration.The influence of the number and layout of floating support and holding stiffness of servo press device on the dynamic response of whirling point was researched to optimize the whirling process and improve the total length quality of large-scale thread further.Finally based on process parapmeters the surface roughness and vibration prediction models were established with regression and BP neural network method which are helpful to guide the pratical production and reduce the waste and improve the economic benefit.